Artificial neural network in breast lesions from fine-needle aspiration cytology smear
Artificial neural networks (ANNs) are applied in engineering and certain medical fields. ANN has immense potential and is rarely been used in breast lesions. In this present study, we attempted to build up a complete robust back propagation ANN model based on cytomorphological data, morphometric dat...
Saved in:
Published in | Diagnostic cytopathology Vol. 42; no. 3; pp. 218 - 224 |
---|---|
Main Authors | , , |
Format | Journal Article |
Language | English |
Published |
United States
Blackwell Publishing Ltd
01.03.2014
Wiley Subscription Services, Inc |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Artificial neural networks (ANNs) are applied in engineering and certain medical fields. ANN has immense potential and is rarely been used in breast lesions. In this present study, we attempted to build up a complete robust back propagation ANN model based on cytomorphological data, morphometric data, nuclear densitometric data, and gray level co‐occurrence matrix (GLCM) of ductal carcinoma and fibroadenomas of breast cases diagnosed on fine‐needle aspiration cytology (FNAC). We selected 52 cases of fibroadenomas and 60 cases of infiltrating ductal carcinoma of breast diagnosed on FNAC by two cytologists. Essential cytological data was quantitated by two independent cytologists (SRM, PD). With the help of Image J software, nuclear morphomeric, densitometric, and GLCM features were measured in all the cases on hematoxylin and eosin‐stained smears. With the available data, an ANN model was built up with the help of Neurointelligence software. The network was designed as 41‐20‐1 (41 input nodes, 20 hidden nodes, 1 output node). The network was trained by the online back propagation algorithm and 500 iterations were done. Learning was adjusted after every iteration. ANN model correctly identified all cases of fibroadenomas and infiltrating carcinomas in the test set. This is one of the first successful composite ANN models of breast carcinomas. This basic model can be used to diagnose the gray zone area of the breast lesions on FNAC. We assume that this model may have far‐reaching implications in future. Diagn. Cytopathol. 2014;42:218–224. © 2013 Wiley Periodicals, Inc. |
---|---|
Bibliography: | istex:FCD2A2FE1A59378552FDE82C2AE2BAFEBDECEED4 ArticleID:DC23026 ark:/67375/WNG-1H2WDV19-0 ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
ISSN: | 8755-1039 1097-0339 |
DOI: | 10.1002/dc.23026 |